Stated beliefs usually have multiple functions, and nowhere is this more obvious than in beliefs related to identify and group membership. The two functions below are not the only two functions but they're important ones, and they sometimes conflict, between an individual's beliefs or even the same belief statement.
There is the obvious, contentful function of a stated belief: saying "I believe X" means the person believes concretely that X, and will make concrete measurable decisions in alignment with that belief. This function of the belief is a truth claim in its most mundane, practical sense. "I believe it's going to snow tomorrow," a person might say, and get out their snow boots and set their alarm a half hour earlier because of the snow. Let's call this the simple content function. A naive and superficial approach to understanding human belief claims assumes that the only thing worth paying attention to is the semantic content.*
But there is another function, which is to reinforce one's own identity, as well as to declare an affiliation with a group of people, or separateness from a group you don't like. Maybe the same person might say "Snow tomorrow or not, global warming is real, and people who accept global warming - like me - are intelligent, informed citizens, and people against it are mindless idiots." This will make the person feel better about him or herself, and will let their social circles know they're "right-thinking". (Note: there's no argument here that people are doing so consciously or intentionally. The crux of the ideas is that people do this automatically.) Let's call this the group affiliation function. It often seems that the majority of stated beliefs are mostly exclamations of group-affiliation.
Stated beliefs are often weighted toward one or the other of these functions, and oftentimes, the more important the group affiliation function is, the less important the simple content is. That is, the more important the group affiliation function is, the less important it is that the belief is coherent, accurate, and actionable. That said, just because a given belief contains a large component of group affiliation, that doesn't necessarily mean the belief is not justified. In the text that follows, I'll refer to contentful beliefs and group affiliation beliefs, but in reality most beliefs have components of both and just lean one direction or the other.
People declare group affiliation in myriad ways (clothing, accent, diet) and sometimes, utterances that aren't even language. Confederates had the Rebel Yell, conquistadors "¡Santiago!" These are the human equivalent of territorial barks and howls. In other cases, "stated beliefs" do us the favor of being clearly incoherent, showing that they're more wordless war-cries than truth claims, not really even being sentences or at least being repeated so much that they're reduced to mantra-like reflections of solidarity and fervor, rather than having semantic content. Examples: Go Chargers! Allahu aqbar! Sieg heil! Are people really commanding football players to run, making a theological claim, or praising victory when they shout those things? No, they're saying "Hooray for my side, in a way that only my side says!"
Interestingly, when people are asked about "beliefs", we're much more likely to think about expansive, identity-defining, high-group-affiliation-component beliefs, rather than mostly mundane and actionable ones. That is to say: someone might tell you "sure, I believe that the big oil companies will keep making money and finding oil, but that hardly rises to the level of a 'belief'. The world was created 6,000 years ago, now I believe that!" But not many of those same people reallocate their retirement accounts when reminded that those oil companies look for oil based on an old-Earth materialist/evolutionary model. That is to say, when you find conflicts between pragmatic/actionable/contentful beliefs and group-affiliation beliefs, people become
Sometimes these two functions of belief statements fail to connect. For instance, many Americans who are passionate about their political ideals are really just passionate about being a member of the liberal or conservative tribe; when they get involved with the actual nuts and bolts of voting or working in a government, suddenly everything seems very mundane and fuzzy and not painted in ideological colors at all. (This gives rise to the old political smartass question of how you pave a street in a Republican way versus a Democratic way.)
Sometimes these two functions of belief statements do connect, and they collide head-on. An interesting exercise is trying to turn group-affiliation beliefs into pragmatic ones by attaching a consequence; a quick-and-dirty way to do this is by asking people to make a bet. If the world really will end on a certain date, I'll loan you money you don't have to pay back until the day after (I tried this one a few years ago with Harold Camping crowd; amazingly, no one took me up on it). If gay marriage will destroy America, you pick the consequence and date it will be realized, and we'll make a bet on it. When you do this to people's group-affiliation beliefs, people either a) change the subject, b) tell you that's offensive (betting or indeed attaching any practical consequence to their tribal-affiliation belief, even though these are often claimed to be the most important beliefs in their life!) or c) come right out and tell you there's really no way to verify or test what they're saying, but insist that what they're saying is still true in some special way or magical other realm (or that they can just tell it's true even if you can't, poor soul). But keep in mind: when you demand verification of an incoherent group-affiliation belief, you're attacking someone's group, and in so doing making a full frontal assault on their identity. Of course they won't like it.**
Sometimes people manage to package a mostly-group-affiliation belief into a complete sentence. Fans holding up their index fingers and shouting "We're Number One!" at a sports upset usually don't believe they're really number one, if you insist on asking them in the moment of their elation. And in fact, we have whole books containing group-affiliation belief statements with the explicit claim that they can't really be verified, but are still somehow true. (Razib Khan excellently describes what theologians do in interpreting these texts as intellectual foam, because the exegesis and even the core text itself usually has very little to do with the religions they catalyzed. The same could be said of most political movements.) Of course, if you're a member of some religious or political ideology and you think I'm talking about you, then assume that I'm talking about those other weird religions or political groups, but do be sure to explain in the comments why your own special club is different and doesn't suffer the same problems.
Group-affiliation is a component of all our claimed beliefs and sometimes (usually?) overwhelms the truth value of those beliefs. And this is not some problem with the influence of Christianity, or modern consumer capitalism, or anything so provincial. It appears quite universal, stemming from the genetics and neurology of humans, and it wouldn't be a matter of education but rather full-on genetic engineering to do away with it. The key is when those incoherent group-affiliation-overloaded beliefs are carried forward by institutions with political power, they often do have effects, and begin impinging on other people, now we have a problem. Case in point: one of my cousins thinks I'm kind of fruity for not caring about pro football, but he doesn't think I'm a bad guy and definitely doesn't consider me a source of family shame for it, or think I should be kept away from children, or anything like that. Another of my cousins thinks that I should have to leave America, or be quiet - interestingly, because he says the majority of America is Catholic (which is even funnier than you think, because our family is Lutherans all the way down.)
You may have noticed the frequent referral to sports fans, because they are a (mostly) harmless manifestation of tribal identification and group-affiliation distorting the truth-claim aspect of not just belief statements, but human cognition itself. You may also have noticed that on average, atheists tend not to be fans of team sports. (I looked for data, but just found other people making the same anecdotal observation.) Why might this be? It may be that there is something cognitively different about people who choose atheism against a background of religion - that as Pascal himself said, we were made in such a way that we cannot believe. We like to pat ourselves on the back and say it's because we're more rational, and without excluding that, it seems very likely that we are atheists in part because we just can't as automatically absorb group-affiliation, through language and otherwise. (Turkish sociologist Fehmi Kaya once implied that atheism was a form of autism. Of course this is incorrect and he later apologized, but you can actually see what he was getting at.) Consequently, in a culture where atheism is not the norm, atheists would also not be likely sports fans. (In China or Scandinavia where atheism is more common and therefore less informative about the person, I would expect there not to be a weaker correlation.) If atheists in general think (and communicate) in terms that more heavily favor the contentful component over the group-affiliation component, then team sports would seem a strange world indeed. It also explains the frustrations that atheists often experience in talking to strongly-identifying religious and political group members.
In closing: it's worth repeating that all these group-affiliation-laden belief statements aren't conscious clever ploys of scoring status points with your favorite in-group. If it was a conscious decision, why would you have a favorite in-group in the first place? They're part of our actual beliefs, to some degree in all of us, and we believe them to be "true". Our need for solidarity dramatically affects our cognition. A recent paper in Psychological Science looked at how extremeness of political beliefs relates to the anchor bias. The anchor bias is simple. If you take two groups of people, and tell group A, "New York and San Francisco are at least 2,000 miles apart; try to guess exactly how far apart they are?" and group B, "They're at least 2,500 miles apart...", guess what? Group A will on average give a smaller number, because they have a smaller anchor. (If you think this is B.S., marketing professionals don't, and they have your money to prove it.) Interestingly, political extremists are less affected by this bias than people closer to the center - they are less influenced by external cues. My prediction is that they would also be less affected by other tests of conformity (the famous Milgram experiment, or standing in elevators facing the back or saying a shape is a different length than it actually is). To be clear that I'm not scoring an own-goal for theists on this one: the point is not that extremism is good, but rather that humans adjust their perception of reality to group norms, and that atheists (in the U.S. anyway) are people who are less likely to do this, and you could likely measure it in these experiments.
*This is a descriptive rather than normative statement. When people state their beliefs, they are mostly not actionable, coherent, accurate beliefs. If robots followed people around trying to translate their belief claims into executable code, most of what they said would have to be non-runnable comment, and the rest would probably kill the person and/or make the world end.
**In 2010 the San Diego Coalition of Reason put up an atheist billboard along an SD freeway. A close friend who is now atheist but was still at the time Christian told me that when she saw it, she experienced it strongly as a personal attack. For atheists who have never so strongly identified as a group I think it's very hard for us to viscerally understand this. This kind of bees-swarming-from-the-hive response is a weapon that organized religion (and any programmed-in-childhood argument-from-authority ideology) has and that we atheists don't have. Consequently any frontal assault should be undertaken only after carefully calculating the risk:benefit.